This study investigates the effects of artificial intelligence (AI) adoption in organizations, specifically within teams. The authors asked the following questions:
- How should a principal optimally deploy limited AI resources to replace workers in a team?
- In a sequential workflow, which workers face the highest risk of AI replacement?
- How does substitution with AI affect both the replaced and non-replaced workers’ wages?
The authors find four main results.
First, firms should not fully replace a single worker every time. Instead, they should randomize. Sometimes a task is done by AI, sometimes by a human. This creates hybrid teams rather than full displacement.
Second, position in the workflow matters. In a three-person team, the middle worker is never optimally replaced because that person keeps information flowing between others. The last worker in the sequence faces the highest risk of replacement. The first worker faces some risk, but less than the last.
Third, firms may choose not to use all available AI capacity. Fully using AI can weaken incentives and force firms to pay higher wages to keep workers motivated.
Fourth, optimal AI adoption increases average wages and reduces wage inequality inside teams.
